Skip Menu |

This queue is for tickets about the Test-Harness CPAN distribution.

Report information
The Basics
Id: 3111
Status: resolved
Priority: 0/
Queue: Test-Harness

People
Owner: andy [...] petdance.com
Requestors: casey [...] geeknest.com
Cc:
AdminCc:

Bug Information
Severity: (no value)
Broken in: (no value)
Fixed in: (no value)



Date: Thu, 31 Jul 2003 19:58:07 -0400
From: Casey West <casey [...] geeknest.com>
To: bug-test-harness [...] rt.cpan.org
Subject: Test::Harness::Straps complains at 100k tests
Sure, that's supposed to guard against recursion and endless loops. But I'm not recursing, so I think I should be able to shut it up via some environment variable at least. What bugs me the most is that it's an arbitrary limit. Casey West -- "Louis Pasteur's theory of germs is ridiculous fiction". -- Pierre Pachet, Professor of Physiology at Toulouse, 1872
[casey@geeknest.com - Thu Jul 31 19:53:23 2003]: Show quoted text
> Sure, that's supposed to guard against recursion and endless loops. > But > I'm not recursing, so I think I should be able to shut it up via some > environment variable at least.
If you specify a plan it won't complain. xoxo, Andy
Subject: Improve 100K warning.
[PETDANCE - Sat Nov 27 14:46:38 2004]: Show quoted text
> [casey@geeknest.com - Thu Jul 31 19:53:23 2003]: >
> > Sure, that's supposed to guard against recursion and endless loops. > > But > > I'm not recursing, so I think I should be able to shut it up via
> some
> > environment variable at least.
> > If you specify a plan it won't complain.
I'm going to guess that at 100,000 tests things are getting a little dynamic and having a plan might not be possible. The flip side is why do you have 100,000 tests in a single file? Anyhow, the check could be made a little smarter. It wasn't put in there to guard against recursion and endless loops. It was put in to guard specificly against this. 1..3 ok 1 ok 2 ok 136211425 See perlbug ID 20020325.002. Basically it prevents TH from allocating a huge array internally. There are better ways of dealing with this. Here's a few starting with the simplest. 1) Up the limit. An empty 1,000,000 element array is maybe a dozen megs. Even 10,000,000 isn't that big. 2) Don't warn if the entry is basically sequental. ie. For every test over N check to see if N-(1..5) exists. Its more than just N-1 because sometimes tests are simply a little bit out of order. 3) Internally, use a hash instead.
Obsoleted in 2.99